(1) Field of the Invention
The invention relates to an image capture device. More specifically, the invention relates to a low-cost three-dimensional digitizer.
(2) Background
There are four broad categories of non-contact three-dimensional digitizers. The first category is referred to as silhouette digitizers because the imaging device repeatedly takes the silhouette of the object as the object is rotated before the imaging device or the imaging device is rotated about the object. This type of digitizer is relatively ineffective at dealing with concavities in a three-dimensional object because the silhouette is unchanged by the concavity.
The second category is timing digitizers. Timing digitizers use a signal source such as a radar source. By determining the amount of time required for the signal to bounce back from the different points on the target object, surface features of the object are revealed. However, such digitizing methods are extremely expensive to implement.
A third category is projected pattern digitizers, in which a pattern of some sort is projected onto the object to be digitized and the dimensions of the object are determined from the interaction of the pattern with the object. Projected pattern digitizers fall into three main subcategories. The subcategories include contour digitizers which use spatial distortion from a projected pattern of contour lines to determine surface features of a three-dimensional object. A next subcategory is interference projected pattern digitizers, which use two sources and then based on the localized interference pattern of the two sources, determine the surface features of the three-dimensional object to be digitized. A third subcategory is referred to as color projected pattern digitizers because this category uses a projected color pattern and resulting color gradients to determine relevant information about the object to be digitized.
A final broad category is stereoscopic digitizers which employ multiple cameras to capture images of the object from different angles. From the picture, such systems perform feature identification. Then a correlation between the features in the different pictures is established to yield three-dimensional data.